Ask HN: Please review our startup: historious
It's at http://historio.us/
I have posted it here once before, and you guys were very helpful, giving a deluge of very good suggestions. Right now I'd like to ask for your opinion on the pricing model.
Our model right now is a free option, which has a limit of 1000 bookmarks for a month, and then it becomes unlimited. This is to avoid people coming in, importing their 20,000 bookmarks and leaving, never to return.
There are also some extra features here: http://historio.us/pricing/
The main problem, and what we'd like your feedback on, is that every bookmark a user adds is expensive, as we have to store the entire page. Right now the service is profitable, but we'd like to improve it so we can scale it better.
What features would make you pay for it? What sort of things would turn you away from it as a free user with a mind to convert later on?
A good point we heard for not having a bookmark limit is that, even with 1,000 bookmarks, users would think twice before bookmarking something, and we don't want that.
Any feedback you could give is greatly appreciated!
It already merges my personal "Web History" (things I search for on Google) with my actual bookmarks (things I click 'bookmark' on a bookmarklet - http://www.google.com/support/chrome/bin/answer.py?hl=en&... ).
The key feature of Google bookmarks is that searching the bookmarks always searches the content of all bookmarked pages and the web history pages. It returns the results with bookmarked pages at the top, and web history below.
It also has tags (they call them labels) and collections (they call them lists), and collections are shareable, and tags are searchable (find X only in Erlang tagged bookmarks).
You can import bookmarks from other services, and export all bookmarks to XML (one of the reasons I stay with Google is that they allow me to move).
My question to you is: What do you offer that trumps that functionality that I already have? Why should I move? Snowman?
A few features that are coming are the ability to import RSS feeds from other places, as well as more social elements (and, of course, support).
For you, though, David, I'll throw in a free subscription!
The killer feature is the cache.
To know that what I bookmark will have a snapshot at that moment in time, citable and impervious to changes that happen to the source.
That is a fantastic feature.
My advice to you would be to highlight that. Bookmarks against things that can and will change, disappear or move are moving targets... what you're offering here is a permanent bookmark as it stood at that moment in time.
That's a big deal, definitely enough to make me consider trying it out. But you need to communicate that this is possible... it would really help researchers and those who use bookmarks as a searchable source over a long period of time (which is when the effects of things changing becomes most obvious).
I think you should possibly look at features like verifying a source (it looks like you cache from the browser, which the user could've modified) and allowing tags to be shared with other users or publicly. Basically... allow academics to use this and to include web citations in their papers and such with histori.us providing the verified cache.
Perhaps even extend the bookmarking functionality to include the ability to highlight and add notes to part of the bookmarked page.
Permanent cache = big deal.
http://buro9.historio.us/search/
You can publish sites there (including tags) and refer people to it by giving them a link of the form http://buro9.historio.us/?q=some+query. Personally, I think that's pretty handy for answering the question "do you have any sites about X?"
No items are made public by default, but do note that, when you do make an item public, you are also sharing the cached version.
Thanks again for the feedback, and I'm good for that free subscription if you need it!
Beware and embrace the slashdotting (insert digg, reddit, HN here).
The cache is both your killer feature and your risk. I hope it's statically stored or can be put in memory quickly... are you using Varnish?
Encourage it's use... if people are bookmarking using histori.us and a site goes down, you've just gained a very large audience to your service. But only if your site survives the onslaught itself.
The biggest problem (by far) is disk usage. The rest of the service is very easily sharded, really, as every user is isolated. Solr is also fantastic (much better than Sphinx, in hindsight we should have gone with that for TP), so that can also scale very well (there's even an implementation of it on hadoop).
We'll add the caching feature to the front page as soon as we finish the current round of A/B testing, thanks again (that feature was basically an afterthought, so it was great that you noted its importance)!
For bonus points, a button to search for the current version of the same page, which may be at a different address from the original one.
http://stavros.historio.us/cached/354988/
Unfortunately, we just lost a day of A/B testing due to an erroneous setting, but we'll start testing this addition to the homepage first thing tomorrow. Thanks again!
This is a just collection of users, perhaps identified strongly by their company email address... and as each user works on a different case they should see their own thing or things of shared tags (multiple people working on the same account).
The company shouldn't lose access to a cached URL just because an employee leaves and decides not to renew their sub with you. Hence the need for a company account.
The company should also be able to control what is made public and to review all public items. This takes care of ensuring that research into defence isn't leaked... perhaps a simple approval process on what is made public (whereby a named person for a tag approves items within that tag being made public).
Then make it so that you bill the legal firm on a monthly basis and show the proportion of storage per tag (they will re-bill their client accounts) and you have a winner.
After that the only thing you have to do is sell it to a few firms.
Tricky pitch for a new company, of course - you haven't been around so long. Any sort of partnership with an existing one would help. Be aware that law firms in general are conservative about technology so it might take a year or two for the idea to catch on. A way around this might be to target law students first. Contact the editors at prominent 10-20 prominent law journals and give them free accounts for the journal and/or themselves as individuals; let them recommend your freemium service to their fellow students. Law librarians are another likely target, they spend quite a bit of time helping people with research tasks.
I think it would help if these 'permacache links' had some distinctive appearance that was easy to type from a printed page, eg 'http://historio.us/citations/username/98734545.htm <- numeric is easier than mixed alphanumeric if you have to copy it by hand.
In fact you could do 10 digits easily using a telephone format, and then ditch the user id string. 'historio.us/citations/####' is easy enough to become a standard link for public citation, and 111-222-3333 gives you room for 10 billion citations before you need to change the naming schema.
As for your second point, due to a security vulnerability with having cached pages on the main domain, we moved them all to another one:
http://cache.historious.net/cached/354992/
I think that is much more readable, and the number doesn't really have to have a fixed number of digits; it can increment infinitely. If we ever need shorter URLs we can base36-encode it and reduce the number of digits right away.
Thanks again for the insight!
Update notifications. Store the page content, url, and date I historified a document. Alert me to any changes in the document; sort of like an RSS feed. Use the wiki format to display changes in the document.
I also like being able to see all my links at once. I am playing around with an idea here[1]. I know there is an "iframe = bad" mentality, but most places use them- more so with the ajax web. If all my links were to the right and there was a blank iframe that I could load them in on the histrio.us website you keep users on your page longer and increase your traffic. Combine this with the updated views and you have a nice little tool.
I would also like some metrics tools- why store it and not doing any thing cool. Maybe the ability to put pages in sets and query those sets for keywords, images, ect. Do something different than google bookmarks/delicious/et al.
Better facebook/twitter integration. Until you get on the social bookmark list and have your icon plastered on pages you should offer this directly. A icon beside an already historified link to facebook or twitter would be enough for a user to know what it is there for. "Jane just historified URL".
Other than that I am content with it just the way it is now. Keep up the good work.
[1] http://christopherwoodall.com/trendy/
Your idea is very very good, and there are many things we can do, especially with document clustering (grouping similar documents), automatically generated keyword clouds, etc etc. We are trying to get the service to a good, stable level for now, and then we'll devote more time to playing with new ideas.
One such idea we're considering is have the search interface be a filter, where the results are continuously decreasing, in real time, as you type more and more keywords.
We already have Twitter integration (as in, posting things to Twitter when you historify them). We also want to add Facebook integration, but its priority is a bit lower right now.
Thanks again, we add all ideas to our wishlist and are very happy to implement most of them (it just might take some time)!
I like the fact that you don't need to sign in with the bookmarklet (contrasted with instapaper).
The incredibly clean interface is another draw for me. A lot of these other sites are all "web 2.0"ish and to me it's bothersome.
I don't really like the chromium extension. I dunno, I think it's kinda useless seeing as though it's off to the side, whereas the bookmarklet is right on top of what I'm viewing. Also I like the notification window of the bookmarklet better than the extension. The window when a "page has already been indexed" on the extension is a bit annoying. I'd like to see maybe a keyboard shortcut for the extension, I think that would make it killer.
A couple things that I would like would be to search through a time range, also better browser integration would be nice. i.e. having pages show up under the autocomplete. The ability to select multiple items would be nice. Having a text only view (a la instapaper, readability). Also I wish there was a shorter url, I always spell historious wrong.
thanks!
It's perfectly fine if you prefer the bookmarklet over the extension, as we spent more time working on it (it affects more users)! We absolutely want to add a shortcut, but don't know how to do it in Chrome yet (some research is in order).
You can already search for time ranges (the syntax is a bit weird, though), it's "added:[2010-07-01T00:00:00Z TO 2010-08-01T00:00:00Z]", for example.
We are also planning to support better autocompletion, as well as bulk edits/deletions. I have never seen the text-only view, but it sounds like a good idea, I'll take a look now! As for the URLs, we also have historius.com and historious.net. Unfortunately we don't have historious.com for symmetry :/
Thanks again!
Sometime when this happen, I don't even go further and never even try the service. Do you you really need to wait for me to click an "activate" link before letting me try your service ?
EDIT: We've changed the signup process to log you in immediately after activating, and we'll remove the activation requirement soon, thank you!
The main advantages I see you having versus Delizzy are:
- Privacy (they require your Delicious username and password)
- Being able to snapshot the page as you read it
The main disadvantages I see you having versus Delizzy are:
- They integrate with Delicious
http://historio.us/
http://historio.us/pricing/
It's a great piece of thinking though - kudos. I've upvoted you both here and there.
This isn't a shot at you specifically because it happens again and again and again and again, but I'm constantly surprised, constantly surprised that people like yourself who have been here for months don't know the rules. It's something you simply need to know, something that you learn from spending time here, and wondering "I wonder why that happens."
I still can't figure out when the downvote arrow appears, though.
http://ycombinator.com/newsfaq.html
>The best part: It's completely free!
It's not completely. I think you've over marketed that bit, would "It's free" do, or even "You can use it free forever" (I know, but it's clear IMO that forever is curtailed by heat-death of the universe, business ending and the like).
I haven't wrapped my mind around your concept completely yet, and I didn't get any hits by searching on your demo, so ... :)
In Opera, you can see the titles, and you're lucky if you can make out the title of the item you're looking for. On historious, you just search for "nginx ubuntu" and it finds the document, even though the title may be "So and so's blog | Webservers".
Another thing is that, with historious, I don't have to think about where to put the bookmark. I just click the button and it's done, I can look at it later or not, it's always going to be there, even if I forget about it. I find myself searching historious first and Google later, as I usually find things that relate to what I want to do now which I had found interesting before but then forgot about.
Do give it a try, it's really changed the way I bookmark, and I love Opera to bits.
EDIT: Not to mention that I can create a custom search where typing "hist <keywords>" will immediately take me to the first result in my historified sites!
It has tags, but nothing social about it, a more nerdy command line type interface, and because its mine :)